how to identify price reversal points - page 9
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
andreybs: Однако готовых качественных индикаторов не обнаружено, конструктивного ответа не получено. Это заставляет задуматься...
It is difficult to judge an indicator from a single picture. Apart from the adjustable level, what parameters does it have? I haven't found USI indicator on the web - is it your development?
The point is that a flat must be associated with a scale (study period). That is, inside the given period the cumulative price movement tends to zero.
And another thing - the given indicator allows to estimate the trend strength but not the presence of the flat. Look, where there should be a flat, the indicator has gone beyond the regulated level.
The indicator is based on moving averages, and has a moving average period setting parameter. Yes, I made the indicator. You are right, the indicator shows strength of trend, shows direction of trend, presence of flat, etc. There are just little tricks, if I may say so, to work with the indicator.
I assume that these tricks are unformalisable.
2 andreybs: looked at your pictures. I see 5 indicators there. Are you sure that all of them are pairwise independent (statistically)?
In the meantime, all the issues have been resolved on their own. The oscillator has been built and the basic filtering has been applied. The forum thread may be closed for lack of use. I have not come across any constructive suggestions here. You can draw conclusions about the local public...
you can draw conclusions about your attitude towards yourself and any other "public".
If you play with your oscillator, re-read the thread, there's something they've tried to tell you
You can draw conclusions about your attitude to yourself and to any other "public".
If you play with your oscillator, read the thread again, there's something I've tried to tell you
"Tried to suggest?" - There is no need to "try". You either have to say something definite or keep quiet. Flooding is no use anyway.
You personally stated that there are no price reversal points, so the question I posed makes no sense. I disagree with that statement. And neither does old man Dow... I'm not going to argue with you about it. There's no point in arguing about it. Especially since you, apparently for lack of argument, immediately start getting personal.
I assume that these tricks are unformalizable.
2 andreybs: I have looked at your drawings. I see 5 indicators there. Are you sure that all of them are pairwise independent (statistically)?
There are only 3 indicators working on the picture - oscillator, zigzag and trend. Yes, they are statistically completely independent and built on different principles.
The indicator is based on moving averages, and has a moving average period setting parameter. Yes, I made the indicator. You are right, the indicator shows strength of trend, shows direction of trend, presence of flat etc. There are just some little tricks, if I may say so, to work with the indicator.
At first I liked the idea of the indicator, then when I tried to reproduce it I faced a problem - at strong price movements during a flat (i.e. at wide horizontal channel) a chart of bearish or bullish component always breaks through the flat setting level. I remembered that I had tried to build an indicator according to the same principle before and faced the same problem. The channel width should be taken into account in the indicator, i.e. the adaptive flat level should be considered. Otherwise, the indicator often deceives. Maybe the eye can see it, but it will not work for MTS, everything should be clear there.
You somewhat misunderstand the point you are making. The essence is as follows. Using historical data you can easily and naturally find the algorithm of finding the reversal points. But the problems appear when you start trading - on future data that you and your TS do not know, your algorithm will not easily and smoothly find these reversal points. Why? - Because the market changes, patterns change. And due to this, the TS can easily and effortlessly fail, especially if you deal with it for 1 hour a week )))). This is the essence of fitting, what they tried to tell you here - everything is easy and effortless on historical data, but not on future data ))))
At first I liked the idea of the indicator, but when I tried to reproduce it, I faced a problem - at strong price movements during the flat (i.e. at wide horizontal channel) the chart of bearish or bullish component always breaks the level of the flat. I remembered that I had tried to build an indicator according to the same principle before and faced the same problem. The channel width should be taken into account in the indicator, i.e. the adaptive flat level should be considered. Otherwise, the indicator often deceives. Maybe the eye can see it, but it will not work for MTS, everything should be clear there.
You are somewhat missing the point of the problem you raised. And the essence is the following. You can always easily and coercively find the algorithm of finding the reversal points on historical data. But the problems appear when you start trading - on future data that you and your TS do not know, your algorithm will not easily and smoothly find these reversal points. Why? - Because the market changes, patterns change. And due to this, the TS can easily and effortlessly fail, especially if you deal with it for 1 hour a week )))). This is the essence of fitting, what they tried to tell you here - everything is easy and effortless on historical data, but not on future data ))))
So you are saying that testing a strategy on history and its behaviour in real life will be totally incompatible. Then why build a strategy tester into the terminal at all?
Just in case - the work of indicators and TS is implemented "honestly", i.e. neither indicator, nor TS "knows" the future quotes at each moment of the simulated time. They operate only with historical data at the moment of the simulated tick. Generally speaking, I believe that the efficiency of such a modeling reaches 90% as it is declared in the terminal. At least it is confirmed by practice. And there are no tricks, just pure mathematics and statistics...
Here are MTS results using the same calculation algorithm as the indicator.
Well, we can't verify that from a picture... :) You don't know what other filters your MTS uses. It is unlikely that it trades on a single indicator.Have you ever had to explain something to a child that is beyond children's comprehension?
You have to admit, it is not an easy task.
The child does not perceive it or perceives it in his or her own way and more and more questions are born in his or her head, which can stump even an adult.
..................
Don't be lazy, read the old threads.
Any invention of any bicycle must be based on knowledge already known to mankind.
And please moderate your self-loathing - it looks unhealthy from the outside.